arXiv preprint 2025: "Towards Online Multi-Modal Social Interaction Understanding" – proposed Online-MMSI-VLM framework for online multimodal social interaction understanding
ACM-MM 2024: "Two in One Go: Single-stage Emotion Recognition with Decoupled Subject-context Transformer" – introduced DSCT for simultaneous subject localization and emotion classification
TMM 2024: "Facial Action Units as a Joint Dataset Training Bridge for Facial Expression Recognition" – developed AU-ViT for improved facial expression recognition via joint training
NeurIPS 2023: "Real3D-AD: A Dataset of Point Cloud Anomaly Detection" – contributed to a new high-resolution 3D point cloud anomaly detection benchmark
ACM-MM 2022: "Rail Detection: An Efficient Row-based Network and a New Benchmark" – proposed Rail-DB dataset and Rail-Net method for railway detection